Ultimate General - Civil War

[Ssnake]

I'm curious how far away we are from really decent procedurally generated maps in some games. Combat Mission had some amazing random maps which were randomly generated and that's going on 20 years old. They wouldn't be good enough for modern games but I'm surprised at the lack of progress along those routes in the time since. Guessing many devs simply aren't looking at that option and investing the resources to further develop it?

 

The question is, who would go at lengths to develop a really good prodecural map generator.

For most popular games - those that are likely to make enough money so that the developer has the cash to finance the development of a really sophisticated procedural generator - huge maps are detrimental to the game experience. About the only guys who need both detailed and huge maps are wargame developers, particularly if their design involves a 3D view. And among the wargame developers that actually made a lot of money with a title, well, there's Combat Mission which was probably the commercially most successful wargame of all time. Like, they probably made 20x more profit than the next best performing wargame. So who, if not the Combat Mission developers could actually do it?

Note that with Shock Force they scaled back the player's task force size from battalion to company team, and that while the 3D artwork looks nicer, the map complexity hasn't increased as far as I can tell.

 

For us, specifically, the customers seem to be happy with pilfering real-life digital terrain databases and then brushing up a few spots by hand. I have a few ideas how we could at least give the map designers a few more tools at hand to prettify maps at least in places that are likely to be visited by human players, but the military isn't very much interested in geotypical terrain that you flood with procedural detail. _I_ would certainly love to do more, but at the end of the day you only have so many man-hours per year to spend on competing development goals that this is somewhere in the middle to lower third of a wish list that is several miles long.

[Harold Jones]

The funny thing is that if I remember correctly, the genesis of Steel Beasts was a mapping program that Al created as part of his post grad work.

[Ssnake]

At least, a software based renderer for rolling terrain from a close-to-the-ground perspective. And then he started looking for an application case and thought, "Why not tanks? How hard can it be?"

[Harold Jones]

That phrase is the software equivalent of 'Kick it in the ass driver you can make it!'

[Ssnake]

In all fairness, I'm probably directly and indirectly responsible for 90% of the scope creep, which constitutes about 99% of all the work ever invested.

[Skywalkre]

 

I'm curious how far away we are from really decent procedurally generated maps in some games. Combat Mission had some amazing random maps which were randomly generated and that's going on 20 years old. They wouldn't be good enough for modern games but I'm surprised at the lack of progress along those routes in the time since. Guessing many devs simply aren't looking at that option and investing the resources to further develop it?

 

The question is, who would go at lengths to develop a really good prodecural map generator.

My first thought - WoT. The maps they put out are generally pretty bad. When you factor in maps that have been reworked multiple times it's something like half of all maps they've released have been pulled. That's a lot of work put in and money spent for nothing. Some of the ones left are still not that great but apparently are left in just for the sake of diversity for the playerbase.

 

Before the 1.0 patch their maps also weren't that pretty or diverse. Playing a WoT game on an old Combat Mission map would have been preferable, gameplay wise, to what players got before the recent rework.

[Ssnake]

But doesn't WoT go with hand-crafted maps?

I thought this was one of their design decisions, that they wanted to have relatively small maps that are generated in a controlled fashion so they can influence the way that these maps can be played, where choke points are, etc; in short, like a classic shooter game where players need to memorize levels if they want to be successful (by kill count metric). After all, WoT is more of an arena fight than it is a wargame.

 

Also, note that - irrespective of their artistic value and excessive slope gradients, the maps are very detailed, and in a way that still looks functionally OK. Maybe they are procedurally pregenerated, but at the end there's still a lot of human effort in the creation of these maps. And even at Belorussian wage levels, the amount of effort that needs to be put into terrain generation is probably the commercially limiting factor for the map sizes.

[Stuart Galbraith]

New mod for this has come out, with the blessing of the devs. Not tried it yet but seems to buff some interesting things.

http://forum.game-labs.net/topic/25749-j-p-rebalance-mod-by-jonnyh13-and-pandakraut-5118/

[Skywalkre]

Played this for the first time in months. One of the filler battles before Antietam was ridiculously hard and it turned me off of the game which I was playing weekly earlier in the year. (I'm all for hard but ridiculous jumps in difficulty for no reason just scream either poor, lazy testing or a complete lack thereof). Managed to beat that scenario today but it was a bloodbath. As a whole these filler battles have turned out to be much nastier than the bigger historical ones.

 

Also learned something new today - your reputation caps at 100. I have no idea how long mine has been up there. Now I'm debating whether it's worth spending some of it or maintaining the morale bonus you get from having it that high (I'm pretty sure this bonus is a big reason my lower quality troops are able to hold against veteran Southern formations as long as they do).

Edited August 23, 2018 by Skywalkre

[Stuart Galbraith]

I really must play it again. I found a strange bug on the Union side where you get minus thousands of points, it has no effect on the game, it just made it annoying. So I dropped that promising lead.

 

It depends what filler battles. For the most part I found them ok. I found Antietam as the Confederacy an absolute bloodbath, though I fancy the AI is easier to beat on the defensive because I had fewer issues playing as the Union side.

 

The Confederacy is rather more fun to play in this one I think. Largely because they are so hard up for men and equipment all the time, if you do badly in a battle you have pretty much shagged yourself for the entire game. The Union has less problems in this regard.

Edited August 23, 2018 by Stuart Galbraith

[Skywalkre]

I think one reason the filler battles are worse ties in to what we talked about months ago - map limitations (plus scenario constraints). The filler battles tend to have no room to maneuver. In the fight I did yesterday it was basically 'push across this open ground to an easily defensible position with almost no time' because there was basically nothing else on the map besides that open ground between myself and the objective. I said screw it and overloaded one side of the map and pushed through what little forest there was and even then my units on the edge of that push barely got within range of the objective before time ran out.

 

On the big historical battle maps I tend to remember having lots of room to maneuver with plenty of time so the above issues didn't come into play.

[Stuart Galbraith]

I tend to use the elite or large units on the smaller scenario's, the logic is they are better able to hold positions, take casualties and their volleys are so overwhelming. But early on when your units are so damn small and poorly trained, it is an issue. The only way round it is to use lots of cheap cannon, they take ages to reload, but when they are firing shrapnel they can easily turn the tide of a battle.

[Skywalkre]

The one interesting thing I noted from yesterday was how (early) rifle units don't seem to necessarily be an upgrade. As I mentioned I tend to avoid going over open terrain unless there is absolutely no other option. I tend to find an approach with some forest cover and push through it. In such terrain, on the offensive, early rifles don't seem all that (and it's easier to get into melee in those scenarios and they're definitely weaker there).

 

Now, on the defensive, they're amazing. I'd love to experiment with a sizeable skirmisher unit equipped with the best long range rifles in the game but at this point (still 1862 I believe) there simply aren't enough in the pipeline to equip even a skirmisher unit.

[Stuart Galbraith]

Id love to have some of those repeater rifles with the regular units, because in real life they seemed insanely powerful. Unfortunately ive never been able to afford them

 

Yes, Im convinced they have modelled the higher muzzle velocity of muskets, at least in damage. You only really notice an advantage in rifles above a couple of hundred yard. It makes a good case for only giving them to really good units, and leavening the rest of divisons out with cheap units equipped with muskets. They can take the hits, then your good old boys can come in and take the credit.

 

Its just a personal view, but I think the AI is much better at attacking than defence. Which is odd, because historically in the US Civil War, it seemed defence was far easier than attack. Maybe its just the AI is better at dumping units on the flank of units already engaged, but that it doesnt work so well in defence, particularly when under artillery. IMHO, downgrading the effect of artillery just a bit would probably help with bloodletting, because I think units are degraded and break too easily under arty fire.

 

Why isnt the rest of tanknet playing this? I would have thought they would all love a musket and rifles game.

[Skywalkre]

Its just a personal view, but I think the AI is much better at attacking than defence. Which is odd, because historically in the US Civil War, it seemed defence was far easier than attack. Maybe its just the AI is better at dumping units on the flank of units already engaged, but that it doesnt work so well in defence, particularly when under artillery. IMHO, downgrading the effect of artillery just a bit would probably help with bloodletting, because I think units are degraded and break too easily under arty fire.

The AI for defense seems to be tied to what were traditional lines of defense in the big historical battles or what the devs feel would be the expected line of defense in the smaller ones. If you deviate from this (by simply going all the way to a flank and rolling up along the line) the AI really struggles. That's what I did a few days ago and that's also what I've done in a few of the historical battles.

 

In the battle a few days ago I was outnumbered 3:2 and had all of one infantry brigade and a skirmisher detachment along a fairly long front which the AI expected me to attack along. Instead I shifted the rest of my infantry and a few guns to the edge of the map and pushed along the forest. Against a human they would have easily recognized this and easily stopped it. Against the AI it left most of its units in the open along the expected avenue of approach (over the open field). Only when my flank reached the next units in line would they turn. Occasionally the AI would send more to help and by trotting out my one full infantry brigade into the field the AI would freak out and reestablish the line.

 

I forget the last major battle I fought but I basically did the same thing. The Southern front I was attacking along was a forest line stretching diagonally across the map. Per the instructions beforehand I was expected to basically pick a point somewhere along this front and attack. Instead, I shifted about 3/4 of my army to a far flank and approached from the side. I then proceeded to roll up the Southerners (who continued to remained deployed as if I would still approach them per the historical outcome).

 

In short, the AI on the defense seems easily beatable with the only real issue being time.

 

On the offense, though, it seems much better as you mention. There appears to be something built in where it can actually read what's going on and react. I know I've played a few battles several times and seen completely different (and sound) approaches by the AI in each fight.

 

 

Why isnt the rest of tanknet playing this? I would have thought they would all love a musket and rifles game.

Good question. It's only $30 on Steam at max price and so far I have 50 hours played and I'm only up to 1862 in the campaign. Even with the recent issue with that overly difficult side mission coupled to AI that's weak on the defense I'd still put this game in the upper 80s out of 100. For the price and quality playing time it promises you you wonder why it's not a bigger hit here and elsewhere.

[Harold Jones]

Good question. It's only $30 on Steam at max price and so far I have 50 hours played and I'm only up to 1862 in the campaign. Even with the recent issue with that overly difficult side mission coupled to AI that's weak on the defense I'd still put this game in the upper 80s out of 100. For the price and quality playing time it promises you you wonder why it's not a bigger hit here and elsewhere.

Why isnt the rest of tanknet playing this? I would have thought they would all love a musket and rifles game.

 

For myself it just doesn't scratch any itches. I can't speak for the other 3 tanknet gamers who aren't playing it.

[Stuart Galbraith]

Somebody has brought out a rebalance mod. Ive not tried it personally, but extending Musket range to 300 feet is probably going to do a lot of units defending.

 

http://forum.game-labs.net/topic/25749-j-p-rebalance-mod-by-jonnyh13-and-pandakraut-81318/

 

 

As for trying, its one of those games at first doesnt seem very appealing. It doesnt LOOK that fantastic compared to say, Total War games. Its only when you actually play it you get what a nice command system it actually has, how it actually uses terrain (unlike some wargames that just do little more than line of sight issues) and when you have an astonishing number of units come at you, you really get some feel of what it must have been like at Isandalwana.

 

Anyway, here is antietam. Still one of the hardest defences ive had to do in a wargame. I felt like Napoleon when I finally beat the Union on this one.

[Harold Jones]

Is there a pause button? I much prefer turn based games.

[Stuart Galbraith]

I think so. There is certainly a half speed button which is as good as.

[Harold Jones]

Thanks, looks interesting enough but I prefer turn based (or at least pausible) games.

[Stuart Galbraith]

Ill have a look tomorrow. Ive not actually played it for a couple of weeks and Ill be honest I never use the pause button so I can state that its got one.

[Skywalkre]

Is there a pause button?

There is.

 

I tend to play by pausing, updating orders, and then using one of the faster speed settings til I reach a point I feel I need to change orders. Almost never play at normal speed.

 

Funny enough the other day when I beat that hard scenario I tried my strat first by playing the entire scenario at one of the faster speeds without ever pausing to give orders. Went back later to 'do it right' and used pause to give updates as I played. I ended up with only like 5% or so fewer casualties (while inflicting fewer losses on the enemy). Was a bit surprised by that.

[Skywalkre]

Played Antietam and it was... surprisingly easy. The win was ridiculously overwhelming (I killed like 3/4 of the Southern forces against maybe 10% losses on my side) and could have likely been worse but I went in blind and was terrified I was pushing too far and walking into a trap in a later phase that just never came. In the end it was a bit of a letdown.

 

I can't help but wonder if there was some bug. The screen before the battle said the enemy army I was facing was over 100k+ in strength but Southern forces in the actual battle were only like 30k (against my 40k). I have no points in Recon so maybe that figure being shown doesn't mean anything. I also vaguely remember reading about the underlying mechanics of how battles are formed months ago. Since my army is fairly small I may be inadvertently abusing whatever these mechanics are and getting more favorable battles.

 

I'm kind of tempted to start up a Southern campaign to get more of a challenge.

[Stuart Galbraith]

Like I say, the AI sucks on the defence. Try it as the Confederates and you are in for some serious hair pulling. It kinda reminds me of the battle of IA Drag for some strange reason.

 

Recon points are a waste I think, till you cant spend on anything else.

[Skywalkre]

So, between YT's algorithm pushing a lot of clips from the Gettysburg movie (and me obliging and watching every one of them) and recently starting Ron Chernow's biography on Grant, I decided to go back and finally finish up my Union campaign. There wasn't anything wrong with the game before. Things were just getting too easy. If only I'd kept playing a little bit further...

 

I'm now two major battles from the end of the campaign. I still haven't lost any fights but there have been some that have been absolutely devastating. I've lost several brigades at this point (and when it's a 3-star veteran unit that really stings) and my strategy of focusing on a more elite but smaller army is coming very close to being unsustainable (I've been converting Reputation>Money after every major battle recently and it's not enough). I used to go into every major battle nearly fully manned. Lately I've been going in at 2/3 strength with some of my Corps (and then barely able to afford replacements after, even when filling up with fresh recruits over veterans to a degree).

 

So, after almost finishing the game here are a bunch of random thoughts on what I've noticed. Keep in mind I have almost no knowledge of the Civil War (I have several books on it way down on my to-read list but I likely won't get to them for months) so I'm curious how what I'm seeing compares to reality:

 

- I just finished Gettysburg. At this point musket-equipped units have become nothing but cannon fodder. Confederate formations apparently all carry rifles at this point. I have had ONE brigade equipped with muskets in the last 4 major battles post more kills than losses. Meanwhile my rifle-equipped brigades are doing 1.5-3x the number of kills to deaths.

 

- As such with the above I stopped reinforcing my musket brigades. Unfortunately I can't afford (even with maxed out econ tech, all full victories, and converting Rep>money) to equip my relatively small Army with all rifles. Pretty soon I'll have nothing left to choose in the tech tree but to enlarge my army. At that point all I'll be able to afford is starting musket-equipped brigades of fresh troops. Considering what the Confederates field they'll be of almost no use (except to die... in droves... and hope they score a few hits to weaken the enemy before my real units push).

 

- On the bright side those fresh brigades will be led by 1-star generals. At this point, with how XP works (and how cheap it is to requisition one), every brigade commander I have is either a 1- or 2-star general (divisional are all 2-star, while only one corps commander has reached 3-star). I know that can't be historically accurate.

 

- Considering I'm near the end of the campaign and close to enlarging my Army I messed around recently with the thought of adding Cavalry or Skirmisher units. Both aren't worth it.

 

- For Cavalry you have to pick armament that either makes them only effective in melee or skirmishers when they dismount. The number of times I've seen a Cavalry charge be viable in this entire campaign I can count on one hand (so that axes the melee route). As skirmishers the AI for them is apparently designed to always fall back which makes them worthless for rushing to a point to hold (and it doesn't help that rifles almost have twice the effective range of the weapons available to Cavalry). In theory they should be decent at scouting but unfortunately some of the time constraints of many battles makes it that you basically have to pick an avenue of attack at the start, based solely on the map and regardless of any enemy positioning (since you don't have it), and go with it. As mentioned earlier in this thread this game really needs a strategic element to the campaign. If you have enough Cavalry in your Army then raids into enemy territory open up as a way to gain resources and Reputation.

 

- Skirmishers are odd because in theory they could be useful if the game gave you enough of the long range rifles to actually equip a single unit. It doesn't. I'm two major battles away from finishing the game and I still can't equip a single Skirmisher unit with enough rifles of greater range than what's available to my Infantry (and annoyingly enough Skirmishers aren't able to buy Infantry rifles... why?). There's zero reason to give them a carbine and have them get decimated by Confederate Infantry carrying rifles (it's what my rifle-equipped units do to Confederate Skirmishers). It's just... bizarre. Why show the weapons available to purchase if you never offer enough to actually equip a unit?

 

- I have no idea what an optimal division makeup is. I'm limited to 4 brigades per division, 4 divisions per corps, and 3 corps max. Right now I'm running a 50/50 mix of Infantry and Artillery per division. It seems to be working fine given the constraints of my Army size (even being light on infantry there's no way I'd give up any artillery). I cheated, sort of, when I came across a thread last year where someone did some testing and discovered 12 guns is the max you should run so that's what I go with. I've also moved completely away from the 12pdr Napoleons (which were upgrades from the 6pdrs I started the campaign with) because with enemy units all sporting rifles now I was losing too many casualties in those units every major battle (and keeping artillery units replenished with vets and lost guns is expensive!). As such, recently, every artillery unit is sporting a rifled gun of some kind (either 10pdr Ordnance or 10pdr Parrot) and stays well in the back (while still getting more kills than the smoothbores were getting). The days of moving up artillery to throw grapeshot on the enemy is no longer viable per where I'm at in the campaign. (Makes me wonder, is this historical? I want to say I read something about this before where by the end of the war artillery crews were getting picked off by rifle fire.)

 

- I found out through some experimentation that Recon is literally worthless. It's bad enough you have to put 2 points in it to get anything out of it. I discovered that enemy Army size varies on your own. I made a save game and put 2 points in and immediately jumped to the next battle. Saw the enemy had an army of 17k. Went back, replenished my troops, and then started up the battle again. Now the enemy had 19k. WTF? If the enemy scales to always be x% your size what's the point in knowing what they start at?

 

- If they ever make another game in this series (it would be FUCKING AMAZING if they made a game with this engine set in the Napoleonic era) they could actually make Recon useful. Since they limit the time you have in these major battles they could have points in Recon translate (eventually) into useful info relayed on the battle map. Don't scale enemy size like now so at low points you just get numbers for the enemy (like you do already). Then at high investment in Recon you'd get more and more info on the battle map of where the enemy is located (handwave it as recon done before the fight). Have lower level points just give a simple 'x brigade in this shaded area' and have maxed out points actually show enemy positioning and numbers just like you see them if you have direct LoS (use a different color so you don't get confused by what you've actually seen).

 

- I've stopped doing minor battles. At this point given my strat of a small, elite Army it just didn't pay economically to fight them. The income wasn't enough to replenish the losses I'd suffer (and these smaller battles tend to be harder than the big ones).

 

- That small, elite approach can be scary in some of these big battles. Early on in Gettysburg I was tasked with holding a part of the line and simply didn't have enough units to spread across it entirely. Had to break off skirmishers from my infantry units to plug the gaps which is... terrifying. Said skirmishers are about 10-15% of your unit and that's the only way you can expand their coverage (no thinning the ranks to double the length of your line like the 20th Maine did at Gettysburg... something that would be nice to add in a future title).

 

All in all it's still an enjoyable game. I'm sitting at 70h and will likely push 80 by the time I'm done. I got this on discount but even with that amount of time put in (and it'll likely be double as I'm going to start a Confederate campaign) it'd be well worth the full price.

Edited May 10, 2019 by Skywalkre